The Hubble constant from the gravitational lens CLASS B0218+357 using the Advanced Camera for Surveys

Mon. Not. R. Astron. Soc. 357, 124–134 (2005) doi:10.1111/j.1365-2966.2004.08618.x The Hubble constant from the gravitational lens CLASS B0218+357 using the Advanced Camera for Surveys T. York,1
N. Jackson,1 I. W. A. Browne,1 O. Wucknitz2 and J. E. Skelton1 † 1 University of Manchester, Jodrell Bank Observatory, Macclesfield, Cheshire SK11 9DL Potsdam, Institut für Physik, Am Neuen Palais 10, 14469 Potsdam, Germany 2 Universität Accepted 2004 November 9. Received 2004 November 4; in original form 2004 May 5 We present deep optical observations of the gravitational lens system CLASS B0218+357, from which we derive an estimate for the Hubble constant (H 0 ). Extensive radio observations using the VLA, MERLIN, the VLBA and VLBI have reduced the degeneracies between H 0 and the mass model parameters in this lens to one involving only the position of the radio-quiet lensing galaxy with respect to the lensed images. B0218+357 has an image separation of only 334 mas, so optical observations have, up until now, been unable to resolve the lens galaxy from the bright lensed images. Using the new Advanced Camera for Surveys (ACS), installed on the Hubble Space Telescope in 2002, we have obtained deep optical images of the lens system and surrounding field. These observations have allowed us to determine the separation between the lens galaxy centre and the brightest image, and so estimate H 0 . We find an optical galaxy position, and hence an H 0 value, that varies depending on our approach to the spiral arms in B0218+357. If the most prominent spiral arms are left unmasked, we find H 0 = 70 ± 5 km s−1 Mpc−1 (95 per cent confidence). If the spiral arms are masked out, we find H 0 = 61 ± 7 km s−1 Mpc−1 (95 per cent confidence). Key words: gravitational lensing – distance scale. 1 INTRODUCTION Objects at cosmological redshifts may be multiply imaged by the action of the gravitational field of foreground galaxies. The first such example of gravitational lensing was the system 0957+561 (Walsh, Carswell & Weymann 1979) in which the core of a background quasar is split into two images 6 arcsec apart. Since then, approximately 70 cases of gravitational lensing by galaxies have been found.1 Refsdal (1964) pointed out that such multiple-image gravitational lens systems could be used to measure the Hubble constant, if the background source was variable, by measuring the time delays between variations of the lensed image and inferring the difference in path-lengths between the corresponding ray paths. The combination of typical deflection angles of ∼1 arcsec around galaxy-mass lens systems with typical cosmological distances implies time delays of the order of weeks, which are in principle readily measurable. Time delays have been measured for 11 gravitational lenses to date: CLASS B0218+357 (Biggs et al. 1999; Cohen et al. 2000), RXJ
E-mail: †Current address: Institute for Astronomy, University of Edinburgh, Royal Observatory, Blackford Hill, Edinburgh EH9 3HJ. 1 A full compilation of known galaxy-mass lens systems is given on the CASTLeS website at http://cfa-www.harvard.edu/glensdata 0911.4+0551 (Hjorth et al. 2002), 0957+561 (Kundic et al. 1997; Oscoz et al. 2001), PG 1115+080 (Schechter et al. 1997), CLASS B1422+231 (Patnaik & Narasimha 2001), SBS 1520+530 (Burud et al. 2002b), CLASS B1600+434 (Koopmans et al. 2000; Burud et al. 2002a), CLASS B1608+656 (Fassnacht et al. 1999, 2002), PKS 1830−211 (Lovell et al. 1998), HE 2149−2745 (Burud et al. 2002a) and HE 1104−1805 (Ofek & Maoz 2003). In principle, given a suitable variable source, the accuracy of the time delay obtained can be better than 5 per cent. This has already been achieved in some cases (e.g. 0957+561) and there is no doubt that diligent future campaigns will further improve accuracy and also produce time delays for more gravitational lens systems. Gravitational lenses provide an excellent prospect of a one-step determination of H 0 on cosmological scales. The major problem is that, in addition to the time delay, a mass model for the lensing galaxy is required in order to determine the shape of the gravitational potential. The model is needed to convert the time delays into angular diameter distances for the lens and source. In double-image lens systems in which the individual images are unresolved, this is a serious problem, as the number of constraints on the mass model (lensed image positions and fluxes) allows no degrees of freedom after the most basic parameters characteristic of the system (source position and flux together with galaxy mass, ellipticity and position angle) have been fitted. In four-image systems the extra constraints provide assistance, and in a few cases, such as the 10-image lens  C 2004 RAS ABSTRACT Hubble constant from CLASS B0218+357 using ACS  C 2004 RAS, MNRAS 357, 124–134 time delays and the best-known galaxy positions are often those with large angular separation, such as 0957+561, and these are the systems in which lensing is most likely to be assisted by a cluster. Again, progress can be made by appropriate modelling of the cluster, and many attempts have been made to do this for 0957+561 (e.g. Kundic et al. 1997; Bernstein & Fischer 1999; Barkana et al. 1999), although there remain uncertainties in the final H 0 estimate. As an alternative, the optical/infrared images of the host galaxy may make an important contribution towards the breaking of degeneracies (Keeton et al. 2000). Kochanek & Schechter (2004) summarize the contribution of lensing to the H 0 debate so far and present options for further progress. One approach is simply to accumulate more H 0 determinations and to rely on statistical arguments to iron out the peculiarities that affect each individual lens system; this approach is vulnerable only to a systematically incorrect understanding of galaxy mass profiles. The alternative approach is to select a lens system in which additional observational effort is most capable of decreasing the systematic errors on the H 0 estimate to acceptable levels. In this paper we argue that CLASS B0218+357 is the best candidate for this process. We describe new HST observations using the Advanced Camera for Surveys (ACS), which are aimed at removing the last major source of systematic uncertainty in this system. We then describe how we use the imaging data to derive a value for the Hubble constant. 2 C L A S S B 0 2 1 8 +3 5 7 A S A K E Y O B J E C T I N H 0 D E T E R M I N AT I O N CLASS B0218+357 was discovered during the early phase of the CLASS survey, the Jodrell Bank–VLA Astrometric Survey (JVAS; Patnaik et al. 1992). B0218+357 consists of two images (A and B) of a background flat-spectrum radio source separated by 0.334 arcsec, together with an Einstein ring (Patnaik et al. 1993). The optical spectrum shows a red continuum source superimposed on a galaxy spectrum. The redshift of the lensing galaxy has been measured optically by Browne et al. (1993) and Stickel (...truncated)